Comment by athenot
9 years ago
This looks really cool. However, boring through rock is the most expensive way to connect 2 places. It only makes sense if the land features or purchase value require it.
Fundamentally, this is about short-circuiting the regular road network and establishing a managed packet network that can bypass congestion. Interestingly, that's also the value proposition of public transit, though it also runs into issues of cost, and too low of a population density make it unfeasible.
I still think we haven't fully leveraged the potential of busses. In most cities, they are slow because they combine the disadvantages of road traffic with the disadvantage of time tables. But what if busses operated on their own dedicated network? Bogota deployed a public transit system made with busses but with a UX of a train[1]. This is a genius idea that could work in many US cities, which have a lot more space to spare.
But back to the original idea, I think we might see in the future a "managed" road network, reserved to self-driving cars which are driven by some central management system, optimising routing for the whole network so as to prevent congestion. This won't require tunnels, just gates, dedicated roads and lots of software.
While I generally agree with what you're saying about embracing bus rapid transit and the cost effectiveness of that option, it's not 100% just a matter of congestion. Surface rail often reduces the value of the land around it due to noise, poor streetscape, etc. whereas subways typically dramatically increase the value of the land around it and don't disrupt the neighborhood. Same goes for large surface roadways.
Also, buses don't have the same capacity as rail does—it's difficult to string together 10 buses and still accelerate quickly, navigate turns, and deal with grade crossings with street vehicles (although guided buses maybe can get partway there? [1]). You can feel the burn in the Bay Area if you try riding the lines on Geary or Mission during hot hours.
And the speeds at which buses run within dense metros often cannot compete with modern subway trains—BART, which is an old system already, hits 70 mph and can go faster, which would be kinda scary to do on city streets in a bus.
[1] https://en.wikipedia.org/wiki/Guided_bus
I'm not convinced that buses don't have the capacity trains do, when you consider that headway between trains is at least 90 seconds. If headway between buses is three seconds, in those 90 seconds, 30 buses can pass by. Assuming a bus has roughly the same capacity as a coach in a train, and that the train has 8 coaches, buses have roughly 4x the capacity.
Not to mention that buses can serve many routes that don't have enough capacity to merit a train, bus routes can be reconfigured without it taking years, and a bus system doesn't cost billions upon billions of dollars to build. Heck, even a single line of a metro system can cost a billion dollars to build.
If buses have dedicated panes where they don't have to interact with normal traffic they can easily hit 70mph. Australian highways in certain places have this system.
My favorite is Vancouver skytrain. While building highways they built a train track on pillars. It's fast and usually got me in and out of the city quicker than a car.
I imagine train tracks on pillars cost effective than underground tunnels. For a dense city it makes sense. Not for outskirts.
> ...boring through rock is the most expensive way to connect 2 places.
I get the impression Musk looked at that expense, brought in some subject matter experts, probed them on why that was the case, and decided there must be a better way. From what I've read of tunnel boring, I can see that happening. Not uncommon for a tunnel boring machine to be custom-built and then abandoned in place at the end of the project. A more general-purpose or modular machine might be what he's thinking about. Unexpected geologic formations come up frequently, a function of the expense of manually-intensive sampling processes today. A robotic sampling crawler that bores hundreds of samples, and automatically classifies as it goes instead of sending to an off-site lab, building a 3D map as it moves, might be what he's looking into.
There might be red tape issues even if these technical challenges are overcome, though. Even if you propose going so deep you avoid all other underground infrastructure altogether, government bureaucracies' permitting and related processes will put a floor on the cost avoidance Musk seeks.
> boring through rock is the most expensive way to connect 2 places
While this is true, getting very good at boring underground could be helpful for building underground habitats on mars.
I never thought of that but it actually makes perfect sense for a Musk venture.
...and mining asteroids.
> But what if busses operated on their own dedicated network?
But if you're building a dedicated network in a major city, why on earth would you limit it to buses? Even with a dedicated network, I don't believe buses can be as fast, comfortable or quiet as rail.
I spent a lot of time in Melbourne, Australia and the trams there are fantastic. You can barely hear them when they're operating so they do a lot to decrease noise, they have a ton of standing room and are quite long, so they can carry a lot of people and I don't believe a bus can come close to the comfort of rail.
Rail fails in a serious earthquake. Large infrastructure can take years to return to operation after disasters.
Even buses are useless during/after disasters in my experience (avoiding legal liability, lack flexibility to mitigate problems, inability to publish changes etc).
Rail fails in a serious earthquake.
Yes, but so do roads. Tunnels especially!
http://i.stuff.co.nz/national/91972268/government-confirms-8...
Maybe because it's cheaper? For example, I don't think India can afford to simultaneously build metro systems that cover all cities, and cover a good enough area of each city. Or, if we can afford it, the govt wouldn't consider spending $100 billion+ the best use of the money. Bus is much more doable.
Trams are more expensive than buses. They have removed tram networks on many cities in Europe in favour of cheap diesel buses but slowly now some are getting back.
Only a visitor so don't know the details, but plenty of Brisbane's tunnels seem to be for buses only...
We have tunnels and a couple of seperate motorways designed soecifically for buses. Which is great for getting from southern suburbs to the city. But anything off the main line can take an extra half hour.
I'm from Bogotá and I can tell you that the bus experience is ugly, it worked for about 10 years but now the buses can't cope with the amount of people Bogotá has.
Yes, it worked for a couple years, but right now it's a chaos. A subway would definitely solve the problem.
I actually think the Viva system in York Region, Ontario, Canada is a great model for this.
Highway 7 started as a car-only multi-lane highway. They took a multi-phase approach to building out a light-rail system. The first phase was adding bus rapid transit, increasing fares, and increasing density of new construction. The second phase was adding dedicated bus lanes and light-rail-ready stations. The final phase will be to add tracks and trains on top of the road.
> However, boring through rock is the most expensive way to connect 2 places.
For now. I think that's the whole point of this company. Reduce cost and risk of drilling just like SpaceX's goal is to reduce cost and risk of space flight.
I wonder why it's so expensive. I agree with musk. It makes sense for humans to have a 3d transport system rather than 2d. The payoff is huge. Moving 200mph through tunnels is the future we should be aiming for.
I also get a feeling he's read "foundation" from Isaac Asimov which talks about this.
I wouldn't want to live on Trantor, though
2 replies →
Regarding the boring expense, an NTBM[1] or subterrene[2], is said to be cheaper and faster.
[1] http://www.sheepletv.com/nuclear-tunnel-boring-machines-swit... [2] https://en.wikipedia.org/wiki/Subterrene
My impression is the boring feasibility and costs will not be the limiter of a very cool tech like this happening, but a more boring practical cost, probably the "zoning" cost required to get compliance and permission for all that state and federal land needed to build the underground link network.
Cheaper? A nuclear machine that melts rock and probably doesn't exist?
I'm not convinced that the video you linked on sheepletv above the "underground alien bases" one is the most reliable info.
Sure, ridiculing such information is easy, especially when you describe it as you have, emphasizing the easiest targets for ridicule. If your intent was to ridicule the information, I get your comment makes sense. But, IMHO, that's not a very substantive contribution for HN. IMHO, it would have been more worthwhile to highlight the less-easy-to-ridicule aspects of the page, such as the quotes from, and discussion about, the Los Alamos patents for NTBM. Please don't interpret this as an attack on you, I'm just trying to rightly do the information justice, encourage substantive discussion. Also, to point out that if your goal is to discover useful information, then it does not work to discredit the valid content of a comment or page, by measuring all its info by the standard of the easiest to ridicule parts.
Finally, from my point of view, getting your ridicule feedback hurts because my aim is to contribute to substantive discussion here, and I don't want to feel discouraged from posting something substantive because it might also have info that could be cherrypicked for ridicule, then letting me feel like I wasted my effort contributing because the info was later unfairly presented. I don't want to feel like I have to do extra work "defending" the original effort to contribute, against such ridicule. So, please, before you rush to ridicule something, really take a look to see if there is substantive content there, and refrain from misrepresenting it, thanks.
5 replies →
Considering the name of the company, maybe they have invented a radical new way to reduce boring costs?
When I am drilling holes for pipework I use some attachment to the power tool that drills one guide hole in the middle and a big pip-sized disk. The material inside that disk is not ground up to pieces, it comes out of the hole whole.
At the moment boring machines do not work like that, a conveyor belt of small pieces comes out. What if A380 fuselage sized pieces of rock came out instead and were conveyed off to the coast for coastal defence purposes, or even land reclamation? Carrying objects weighing thousands of tonnes to the sea is not simple but the amount of material excavated the hard way - ground to pieces first - would be significantly lower.
Maybe they have a 'coring machine' rather than a 'boring machine'?
1 reply →
Also, if you accept the assumption that energy will get cheaper in the future, some boring project may become financially feasible.
It's quite possible they could reduce the cost if they found a way to pipe the excavated rock into surrounding regions of less dense rock, through pipes drilled at intervals along the tunnel. Meaning the excavated rock never needs to go up to the surface, it can just be digested and fed as a slurry into neighbouring areas.
He has said he thinks he can reduce the costs by a factor of ten, but except for having spare parts on hand to reduce downtime I don't think he has been specific about his ideas. He bought a conventional boring machine recently (I don't know if it has been delivered yet) to get more data.
The boring could theoretically be done with close to zero energy, and the cost of the tunnel is just to remove the material. Breaking the chemical bonds along the surface of the tunnel is actually a pretty tiny amount of energy, even if it's solid rock.
Ride the relatively new Silver Line in Boston to see what that's like. Hint: It's not good.